Community Graphs compliment Social Graphs by mapping our relationships within the context of groups.

• Members of any community are able to give positive attestations to the extent of other people’s involvement.
  ‍
• Open and composable attestation algorithms can be used to calculate membership levels for each person.
  ‍
• Community Graphs unlocks new possibilities in community stewardship, governance, identity verification, and reputation systems.
  ‍

Attestations are positive, private, subjective, and non-binary endorsements of people’s involvement in groups.

• ‍Positive: Attestations in Community Graphs are always positive endorsements, as categorizing members by centrality or periphery in a group is a sensitive matter. Positive attestations focus on affirming contributions without creating a divisive or judgmental environment.
  ‍
• Private:  The privacy of attestations is crucial to avoid issues inherent in public vouching systems, such as social pressures, collusion, bribery, and extortion. In public systems, individuals may feel compelled to give high scores to maintain friendships or avoid conflict. They could also engage in reciprocal arrangements or unethical exchanges. By making attestations private, Community Graphs prevent these behaviors while maintaining the integrity of how members recognize contributions.
  ‍
• ‍Subjective: The subjectivity of attestations in Community Graphs highlights the unique and varied ways people recognize and value contributions within a group. Unlike rigid, formula-based systems, subjective attestations reflect individual perspectives, making the system more resilient and harder to manipulate. For example, one person might value leadership, while another might appreciate efficiency or effort, leading to diverse reasons for acknowledging someone's contributions. This inherent variability in how members attest to each other’s involvement strengthens the system’s robustness and resistance to gaming.
  ‍
• ‍Non-binary: Attestations are distributed proportionally, not equally, to reflect a more nuanced view of membership. This approach captures the depth of involvement beyond a simple "in or out" status, allowing for a more accurate representation of each person’s contributions within a group. The attestation algorithm ensures that membership is contextual, providing a richer understanding of an individual's role.
  ‍

Giving attestations are easy!

• First, sign up for CG with the email associated with your community. Follow the instructions to get set up.
• Find and click on your community from the main discover list. Click the “Join” button at the bottom of the screen.
• Then, open the “Community Members” page.

• At the top you will see a list of all the current members that have registered and joined this group
• The attestation section is at the bottom of the screen (you may need to maximize it)

• Then, you can drag peoples’ avatars from the member list and drop them anywhere on the line in the attestation section

• The farther to the right you place anyone, the higher attestation you are giving them. So if you think someone is very core to a group, place them far to the right, where left equals lower attestations, but still positive endorsements!
• If you don’t know someone or don’t think they are a member of a group, simply don’t attest to their membership in that group
  ‍

Membership levels in Community Graphs reflect the depth of each member’s involvement, based on the attestations given by other members. These levels can be calculated using various open and customizable attestation algorithms. Communities have the flexibility to apply specific methods, like quadratic formulas, to capture a more nuanced understanding of participation. However, these algorithms are optional—communities can choose how to calculate membership levels based on their own needs, making the system adaptable to different contexts and priorities.
‍

Understanding the extent of people’s involvement in groups unlocks a whole new range of tools that can be built to improve group alignment and trust.

• Sybil resistance: Imagine if Bob belongs to several offline communities where people meet in person. If members in those communities attest to Bob being a human being having a unique identity, and that identity holds across several of those communities, I can have a high level of trust that Bob’s identity, even when used online, is not a sybil, even if I am not a member of one of those communities.
  ‍
• Identity verification: By analyzing attestations across multiple communities, Community Graphs help verify a person’s identity. For example, if different groups endorse a person’s involvement, their identity can be validated as authentic, making it harder for impersonators to deceive communities.
  ‍
• Community stewardship: Membership levels help identify individuals who consistently contribute to a group’s growth and welfare. These individuals can be trusted with greater responsibilities, such as facilitating discussions or managing resources.
  ‍
• Governance: Community Graphs enable more nuanced voting systems, where members’ voting power corresponds to their level of engagement. This ensures decisions are driven by those most invested in the community's success, creating a fairer and more effective governance structure.
  ‍
• Reputation systems: The intuition around this comes from how our offline social reputations work. We build trust and credibility not just with individuals, but within the context of groups and communities. And then that carries to other communities through overlapping memberships.
  ‍

Community Graphs is open and modular, so you can build something unique on top of it that enhances group dynamics, governance, trust, or reputation systems.

We are at an early stage of development and are expanding in functionality quickly! Please let us know if there are any features that can aid you in creating your vision with Community Graphs.

Access the Community Graphs Public API document here
‍

If you need any help, have any feedback, or want to build with us, please do not hesitate to reach out or share in our Telegram channel
‍